Are CO₂ Emissions a Health Hazard or Just a Side Effect of Progress?#

With the data prepared, we now present three visualizations supporting each perspective. Each visualization is accompanied by a brief insight and an explanation linking it to the respective argument.

Perspective 1:#

Do Higher CO₂ Emissions Shorten Healthy Lifespans? This perspective expects to see negative correlations or warning signs that pollution from CO₂ emissions adversely affects health. We look for evidence that countries with higher emissions have lower healthy life expectancy, or that rapid emission growth is associated with stagnating health outcomes.

Visualization 1.#

CO₂ per Capita vs Healthy Life Expectancy (2019) – Is higher carbon output linked to lower healthy life expectancy?

Hide code cell source
# ── Complete Interactieve Grouped Barplot voor 2019 ──

import pandas as pd
import numpy as np
import plotly.graph_objects as go
import plotly.io as pio

# Renderer instellen voor inline weergave
pio.renderers.default = "notebook_connected"

# 1) Data inladen
co2 = pd.read_csv('owid-co2-data.csv', usecols=['country','year','co2_per_capita'])
hale = pd.read_csv('C64284D_ALL_LATEST.csv')

# 2) Healthy life expectancy filter & renamen
hale = (
    hale.loc[
        (hale['IND_NAME']=='Healthy life expectancy (at birth)') &
        (hale['DIM_SEX']=='TOTAL'),
        ['GEO_NAME_SHORT','DIM_TIME','AMOUNT_N']
    ]
    .rename(columns={
        'GEO_NAME_SHORT':'country',
        'DIM_TIME':'year',
        'AMOUNT_N':'healthy_life_expectancy'
    })
)

# 3) Jaar-kolommen als int
co2['year']  = co2['year'].astype(int)
hale['year'] = hale['year'].astype(int)

# 4) Harmoniseer landnaam en merge voor 2019
co2['country'] = co2['country'].replace({'United States':'United States of America'})
merged = pd.merge(co2, hale, on=['country','year'], how='inner')
df2019 = merged[merged['year']==2019].dropna(subset=['co2_per_capita','healthy_life_expectancy'])

# 5) Specifieke landen in juiste volgorde
countries = [
    'United States of America',
    'France',
    'India'
]
df3 = df2019.set_index('country').loc[countries]

# 6) X-positie en barbreedte
x = np.arange(len(countries))
width = 0.35

# 7) Bouw de interactieve figuur
fig = go.Figure()

# CO₂ per Capita (linker y-as)
fig.add_trace(go.Bar(
    x=x - width/2,
    y=df3['co2_per_capita'],
    name='CO₂ per Capita (tons)',
    marker_color='orange',
    marker_line_width=1,
    marker_line_color='black',
    hovertemplate='CO₂: %{y:.2f} ton<extra></extra>',
    yaxis='y1'
))

# Healthy Life Expectancy (rechter y-as)
fig.add_trace(go.Bar(
    x=x + width/2,
    y=df3['healthy_life_expectancy'],
    name='Healthy Life Expectancy (years)',
    marker_color='blue',
    marker_line_width=1,
    marker_line_color='black',
    hovertemplate='Gezonde levensverwachting: %{y:.1f} jaar<extra></extra>',
    yaxis='y2'
))

# 8) Layout met dubbele y-as en styling
fig.update_layout(
    title="Emissions vs Healthy Life Expectancy: U.S. vs France vs India (2019)",
    xaxis=dict(
        tickmode='array',
        tickvals=x,
        ticktext=countries,
        tickangle=15,
        title=dict(text='Country')
    ),
    yaxis=dict(
        title=dict(text='CO₂ per Capita (tons)', font=dict(color='orange')),
        tickfont=dict(color='orange')
    ),
    yaxis2=dict(
        title=dict(text='Healthy Life Expectancy (years)', font=dict(color='blue')),
        tickfont=dict(color='blue'),
        overlaying='y',
        side='right'
    ),
    legend=dict(
        orientation='h',
        yanchor='bottom',
        y=1.02,
        x=0.5,
        xanchor='center'
    ),
    bargap=0.2,
    margin=dict(t=80, b=50, l=60, r=60)
)

# 9) Toon de interactieve grafiek
fig.show()

# (Optioneel) Exporteer als standalone HTML
# fig.write_html("emissions_health_2019.html", include_plotlyjs='cdn')

Caption: Each point is a country in 2019. We might expect a downward trend if emissions were broadly harming health, but the pattern is not clear-cut. Many high-emission countries (far right) actually enjoy high healthy life expectancy, while low-emission countries (far left) often have low healthy life expectancy. The fitted trendline (from an OLS regression) is nearly flat, indicating no strong linear relationship globally.

Insight: At a global level, there is no obvious negative correlation between CO₂ emissions per person and healthy life expectancy. For example, the United States and Gulf countries have some of the highest per-capita CO₂ emissions yet still report healthy life expectancies around 65–70+ years. In contrast, countries with minimal emissions (mostly low-income nations) cluster in the lower-left, with healthy life expectancy often below 60 years. This suggests that factors other than emissions (like economic development) are dominating the health outcomes. However, supporters of Perspective 1 point out that this global view could mask specific health costs of emissions for instance, chronic air pollution in rapidly industrializing countries might be limiting further health gains. The lack of a clear inverse trend here hints that CO₂’s impact on health is indirect and tangled with development rather than a simple one-to-one effect.

Visualization 2. Emissions and Health in Selected Countries (2019)#

– Case comparison of a high emitter vs. a low emitter vs. a moderate emitter.

Hide code cell source
import pandas as pd
import plotly.graph_objects as go
import numpy as np

# 1) Data inladen en voorbereiden
co2 = pd.read_csv('owid-co2-data.csv', usecols=['country','year','co2_per_capita'])
hale = pd.read_csv('C64284D_ALL_LATEST.csv')

# Harmoniseer landnaam
co2['country'] = co2['country'].replace({'United States':'United States of America'})

# Filter Healthy life expectancy
hale = (
    hale.loc[
        (hale['IND_NAME']=='Healthy life expectancy (at birth)') &
        (hale['DIM_SEX']=='TOTAL'),
        ['GEO_NAME_SHORT','DIM_TIME','AMOUNT_N']
    ]
    .rename(columns={
        'GEO_NAME_SHORT':'country',
        'DIM_TIME':'year',
        'AMOUNT_N':'healthy_life_expectancy'
    })
)

# Jaar als int, merge en filter voor 2019 + geselecteerde landen
co2['year'] = co2['year'].astype(int)
hale['year'] = hale['year'].astype(int)
merged = pd.merge(co2, hale, on=['country','year'], how='inner')

countries = [
    'United States of America',
    'France',
    'India'
]

df2019 = (
    merged
    .query("year == 2019 and country in @countries")
    .set_index('country')
    .loc[countries]
    .reset_index()
)

# 2) X-waarden en barbreedte
x = np.arange(len(countries))
width = 0.4

# 3) Bouw de interactieve figuur
fig = go.Figure()

# CO₂ per Capita (linker y-as)
fig.add_trace(go.Bar(
    x=x - width/2,
    y=df2019['co2_per_capita'],
    name='CO₂ per Capita (tons)',
    marker_color='orange',
    yaxis='y1',
    hovertemplate='CO₂: %{y} ton<extra></extra>'
))

# Healthy Life Expectancy (rechter y-as)
fig.add_trace(go.Bar(
    x=x + width/2,
    y=df2019['healthy_life_expectancy'],
    name='Healthy Life Expectancy (years)',
    marker_color='blue',
    yaxis='y2',
    hovertemplate='Gezonde levensverwachting: %{y} jaar<extra></extra>'
))

# 4) Layout met dubbele y-as en styling gelijk aan matplotlib
fig.update_layout(
    title="Emissions vs Healthy Life Expectancy: U.S. vs France vs India (2019)",
    xaxis=dict(
        tickmode='array',
        tickvals=x,
        ticktext=countries,
        tickangle=15,
        title='Country'
    ),
    yaxis=dict(
        title='CO₂ per Capita (tons)',
        titlefont=dict(color='orange'),
        tickfont=dict(color='orange')
    ),
    yaxis2=dict(
        title='Healthy Life Expectancy (years)',
        titlefont=dict(color='blue'),
        tickfont=dict(color='blue'),
        overlaying='y',
        side='right'
    ),
    legend=dict(
        orientation='h',
        yanchor='bottom',
        y=1.02,
        x=0.5,
        xanchor='center'
    ),
    bargap=0.2,
    margin=dict(t=80, b=50, l=60, r=60)
)

# 5) Toon interactieve grafiek
fig.show()

# Optioneel: exporteer naar standalone HTML
# fig.write_html("emissions_health_2019.html", include_plotlyjs='cdn')
---------------------------------------------------------------------------
ValueError                                Traceback (most recent call last)
Cell In[2], line 73
     63 fig.add_trace(go.Bar(
     64     x=x + width/2,
     65     y=df2019['healthy_life_expectancy'],
   (...)     69     hovertemplate='Gezonde levensverwachting: %{y} jaar<extra></extra>'
     70 ))
     72 # 4) Layout met dubbele y-as en styling gelijk aan matplotlib
---> 73 fig.update_layout(
     74     title="Emissions vs Healthy Life Expectancy: U.S. vs France vs India (2019)",
     75     xaxis=dict(
     76         tickmode='array',
     77         tickvals=x,
     78         ticktext=countries,
     79         tickangle=15,
     80         title='Country'
     81     ),
     82     yaxis=dict(
     83         title='CO₂ per Capita (tons)',
     84         titlefont=dict(color='orange'),
     85         tickfont=dict(color='orange')
     86     ),
     87     yaxis2=dict(
     88         title='Healthy Life Expectancy (years)',
     89         titlefont=dict(color='blue'),
     90         tickfont=dict(color='blue'),
     91         overlaying='y',
     92         side='right'
     93     ),
     94     legend=dict(
     95         orientation='h',
     96         yanchor='bottom',
     97         y=1.02,
     98         x=0.5,
     99         xanchor='center'
    100     ),
    101     bargap=0.2,
    102     margin=dict(t=80, b=50, l=60, r=60)
    103 )
    105 # 5) Toon interactieve grafiek
    106 fig.show()

File ~\AppData\Local\Programs\Python\Python313\Lib\site-packages\plotly\graph_objs\_figure.py:219, in Figure.update_layout(self, dict1, overwrite, **kwargs)
    193 def update_layout(self, dict1=None, overwrite=False, **kwargs) -> "Figure":
    194     """
    195 
    196     Update the properties of the figure's layout with a dict and/or with
   (...)    217 
    218     """
--> 219     return super().update_layout(dict1, overwrite, **kwargs)

File ~\AppData\Local\Programs\Python\Python313\Lib\site-packages\plotly\basedatatypes.py:1413, in BaseFigure.update_layout(self, dict1, overwrite, **kwargs)
   1389 def update_layout(self, dict1=None, overwrite=False, **kwargs):
   1390     """
   1391     Update the properties of the figure's layout with a dict and/or with
   1392     keyword arguments.
   (...)   1411         The Figure object that the update_layout method was called on
   1412     """
-> 1413     self.layout.update(dict1, overwrite=overwrite, **kwargs)
   1414     return self

File ~\AppData\Local\Programs\Python\Python313\Lib\site-packages\plotly\basedatatypes.py:5215, in BasePlotlyType.update(self, dict1, overwrite, **kwargs)
   5213     with self.figure.batch_update():
   5214         BaseFigure._perform_update(self, dict1, overwrite=overwrite)
-> 5215         BaseFigure._perform_update(self, kwargs, overwrite=overwrite)
   5216 else:
   5217     BaseFigure._perform_update(self, dict1, overwrite=overwrite)

File ~\AppData\Local\Programs\Python\Python313\Lib\site-packages\plotly\basedatatypes.py:3989, in BaseFigure._perform_update(plotly_obj, update_obj, overwrite)
   3983 validator = plotly_obj._get_prop_validator(key)
   3985 if isinstance(validator, CompoundValidator) and isinstance(val, dict):
   3986 
   3987     # Update compound objects recursively
   3988     # plotly_obj[key].update(val)
-> 3989     BaseFigure._perform_update(plotly_obj[key], val)
   3990 elif isinstance(validator, CompoundArrayValidator):
   3991     if plotly_obj[key]:
   3992         # plotly_obj has an existing non-empty array for key
   3993         # In this case we merge val into the existing elements

File ~\AppData\Local\Programs\Python\Python313\Lib\site-packages\plotly\basedatatypes.py:3966, in BaseFigure._perform_update(plotly_obj, update_obj, overwrite)
   3964     err = _check_path_in_prop_tree(plotly_obj, key, error_cast=ValueError)
   3965     if err is not None:
-> 3966         raise err
   3968 # Convert update_obj to dict
   3969 # --------------------------
   3970 if isinstance(update_obj, BasePlotlyType):

ValueError: Invalid property specified for object of type plotly.graph_objs.layout.YAxis: 'titlefont'

Did you mean "tickfont"?

    Valid properties:
        anchor
            If set to an opposite-letter axis id (e.g. `x2`, `y`),
            this axis is bound to the corresponding opposite-letter
            axis. If set to "free", this axis' position is
            determined by `position`.
        automargin
            Determines whether long tick labels automatically grow
            the figure margins.
        autorange
            Determines whether or not the range of this axis is
            computed in relation to the input data. See `rangemode`
            for more info. If `range` is provided and it has a
            value for both the lower and upper bound, `autorange`
            is set to False. Using "min" applies autorange only to
            set the minimum. Using "max" applies autorange only to
            set the maximum. Using *min reversed* applies autorange
            only to set the minimum on a reversed axis. Using *max
            reversed* applies autorange only to set the maximum on
            a reversed axis. Using "reversed" applies autorange on
            both ends and reverses the axis direction.
        autorangeoptions
            :class:`plotly.graph_objects.layout.yaxis.Autorangeopti
            ons` instance or dict with compatible properties
        autoshift
            Automatically reposition the axis to avoid overlap with
            other axes with the same `overlaying` value. This
            repositioning will account for any `shift` amount
            applied to other axes on the same side with `autoshift`
            is set to true. Only has an effect if `anchor` is set
            to "free".
        autotickangles
            When `tickangle` is set to "auto", it will be set to
            the first angle in this array that is large enough to
            prevent label overlap.
        autotypenumbers
            Using "strict" a numeric string in trace data is not
            converted to a number. Using *convert types* a numeric
            string in trace data may be treated as a number during
            automatic axis `type` detection. Defaults to
            layout.autotypenumbers.
        calendar
            Sets the calendar system to use for `range` and `tick0`
            if this is a date axis. This does not set the calendar
            for interpreting data on this axis, that's specified in
            the trace or via the global `layout.calendar`
        categoryarray
            Sets the order in which categories on this axis appear.
            Only has an effect if `categoryorder` is set to
            "array". Used with `categoryorder`.
        categoryarraysrc
            Sets the source reference on Chart Studio Cloud for
            `categoryarray`.
        categoryorder
            Specifies the ordering logic for the case of
            categorical variables. By default, plotly uses "trace",
            which specifies the order that is present in the data
            supplied. Set `categoryorder` to *category ascending*
            or *category descending* if order should be determined
            by the alphanumerical order of the category names. Set
            `categoryorder` to "array" to derive the ordering from
            the attribute `categoryarray`. If a category is not
            found in the `categoryarray` array, the sorting
            behavior for that attribute will be identical to the
            "trace" mode. The unspecified categories will follow
            the categories in `categoryarray`. Set `categoryorder`
            to *total ascending* or *total descending* if order
            should be determined by the numerical order of the
            values. Similarly, the order can be determined by the
            min, max, sum, mean, geometric mean or median of all
            the values.
        color
            Sets default for all colors associated with this axis
            all at once: line, font, tick, and grid colors. Grid
            color is lightened by blending this with the plot
            background Individual pieces can override this.
        constrain
            If this axis needs to be compressed (either due to its
            own `scaleanchor` and `scaleratio` or those of the
            other axis), determines how that happens: by increasing
            the "range", or by decreasing the "domain". Default is
            "domain" for axes containing image traces, "range"
            otherwise.
        constraintoward
            If this axis needs to be compressed (either due to its
            own `scaleanchor` and `scaleratio` or those of the
            other axis), determines which direction we push the
            originally specified plot area. Options are "left",
            "center" (default), and "right" for x axes, and "top",
            "middle" (default), and "bottom" for y axes.
        dividercolor
            Sets the color of the dividers Only has an effect on
            "multicategory" axes.
        dividerwidth
            Sets the width (in px) of the dividers Only has an
            effect on "multicategory" axes.
        domain
            Sets the domain of this axis (in plot fraction).
        dtick
            Sets the step in-between ticks on this axis. Use with
            `tick0`. Must be a positive number, or special strings
            available to "log" and "date" axes. If the axis `type`
            is "log", then ticks are set every 10^(n*dtick) where n
            is the tick number. For example, to set a tick mark at
            1, 10, 100, 1000, ... set dtick to 1. To set tick marks
            at 1, 100, 10000, ... set dtick to 2. To set tick marks
            at 1, 5, 25, 125, 625, 3125, ... set dtick to
            log_10(5), or 0.69897000433. "log" has several special
            values; "L<f>", where `f` is a positive number, gives
            ticks linearly spaced in value (but not position). For
            example `tick0` = 0.1, `dtick` = "L0.5" will put ticks
            at 0.1, 0.6, 1.1, 1.6 etc. To show powers of 10 plus
            small digits between, use "D1" (all digits) or "D2"
            (only 2 and 5). `tick0` is ignored for "D1" and "D2".
            If the axis `type` is "date", then you must convert the
            time to milliseconds. For example, to set the interval
            between ticks to one day, set `dtick` to 86400000.0.
            "date" also has special values "M<n>" gives ticks
            spaced by a number of months. `n` must be a positive
            integer. To set ticks on the 15th of every third month,
            set `tick0` to "2000-01-15" and `dtick` to "M3". To set
            ticks every 4 years, set `dtick` to "M48"
        exponentformat
            Determines a formatting rule for the tick exponents.
            For example, consider the number 1,000,000,000. If
            "none", it appears as 1,000,000,000. If "e", 1e+9. If
            "E", 1E+9. If "power", 1x10^9 (with 9 in a super
            script). If "SI", 1G. If "B", 1B.
        fixedrange
            Determines whether or not this axis is zoom-able. If
            true, then zoom is disabled.
        gridcolor
            Sets the color of the grid lines.
        griddash
            Sets the dash style of lines. Set to a dash type string
            ("solid", "dot", "dash", "longdash", "dashdot", or
            "longdashdot") or a dash length list in px (eg
            "5px,10px,2px,2px").
        gridwidth
            Sets the width (in px) of the grid lines.
        hoverformat
            Sets the hover text formatting rule using d3 formatting
            mini-languages which are very similar to those in
            Python. For numbers, see:
            https://github.com/d3/d3-format/tree/v1.4.5#d3-format.
            And for dates see: https://github.com/d3/d3-time-
            format/tree/v2.2.3#locale_format. We add two items to
            d3's date formatter: "%h" for half of the year as a
            decimal number as well as "%{n}f" for fractional
            seconds with n digits. For example, *2016-10-13
            09:15:23.456* with tickformat "%H~%M~%S.%2f" would
            display "09~15~23.46"
        insiderange
            Could be used to set the desired inside range of this
            axis (excluding the labels) when `ticklabelposition` of
            the anchored axis has "inside". Not implemented for
            axes with `type` "log". This would be ignored when
            `range` is provided.
        labelalias
            Replacement text for specific tick or hover labels. For
            example using {US: 'USA', CA: 'Canada'} changes US to
            USA and CA to Canada. The labels we would have shown
            must match the keys exactly, after adding any
            tickprefix or ticksuffix. For negative numbers the
            minus sign symbol used (U+2212) is wider than the
            regular ascii dash. That means you need to use 1
            instead of -1. labelalias can be used with any axis
            type, and both keys (if needed) and values (if desired)
            can include html-like tags or MathJax.
        layer
            Sets the layer on which this axis is displayed. If
            *above traces*, this axis is displayed above all the
            subplot's traces If *below traces*, this axis is
            displayed below all the subplot's traces, but above the
            grid lines. Useful when used together with scatter-like
            traces with `cliponaxis` set to False to show markers
            and/or text nodes above this axis.
        linecolor
            Sets the axis line color.
        linewidth
            Sets the width (in px) of the axis line.
        matches
            If set to another axis id (e.g. `x2`, `y`), the range
            of this axis will match the range of the corresponding
            axis in data-coordinates space. Moreover, matching axes
            share auto-range values, category lists and histogram
            auto-bins. Note that setting axes simultaneously in
            both a `scaleanchor` and a `matches` constraint is
            currently forbidden. Moreover, note that matching axes
            must have the same `type`.
        maxallowed
            Determines the maximum range of this axis.
        minallowed
            Determines the minimum range of this axis.
        minexponent
            Hide SI prefix for 10^n if |n| is below this number.
            This only has an effect when `tickformat` is "SI" or
            "B".
        minor
            :class:`plotly.graph_objects.layout.yaxis.Minor`
            instance or dict with compatible properties
        mirror
            Determines if the axis lines or/and ticks are mirrored
            to the opposite side of the plotting area. If True, the
            axis lines are mirrored. If "ticks", the axis lines and
            ticks are mirrored. If False, mirroring is disable. If
            "all", axis lines are mirrored on all shared-axes
            subplots. If "allticks", axis lines and ticks are
            mirrored on all shared-axes subplots.
        nticks
            Specifies the maximum number of ticks for the
            particular axis. The actual number of ticks will be
            chosen automatically to be less than or equal to
            `nticks`. Has an effect only if `tickmode` is set to
            "auto".
        overlaying
            If set a same-letter axis id, this axis is overlaid on
            top of the corresponding same-letter axis, with traces
            and axes visible for both axes. If False, this axis
            does not overlay any same-letter axes. In this case,
            for axes with overlapping domains only the highest-
            numbered axis will be visible.
        position
            Sets the position of this axis in the plotting space
            (in normalized coordinates). Only has an effect if
            `anchor` is set to "free".
        range
            Sets the range of this axis. If the axis `type` is
            "log", then you must take the log of your desired range
            (e.g. to set the range from 1 to 100, set the range
            from 0 to 2). If the axis `type` is "date", it should
            be date strings, like date data, though Date objects
            and unix milliseconds will be accepted and converted to
            strings. If the axis `type` is "category", it should be
            numbers, using the scale where each category is
            assigned a serial number from zero in the order it
            appears. Leaving either or both elements `null` impacts
            the default `autorange`.
        rangebreaks
            A tuple of
            :class:`plotly.graph_objects.layout.yaxis.Rangebreak`
            instances or dicts with compatible properties
        rangebreakdefaults
            When used in a template (as
            layout.template.layout.yaxis.rangebreakdefaults), sets
            the default property values to use for elements of
            layout.yaxis.rangebreaks
        rangemode
            If "normal", the range is computed in relation to the
            extrema of the input data. If "tozero", the range
            extends to 0, regardless of the input data If
            "nonnegative", the range is non-negative, regardless of
            the input data. Applies only to linear axes.
        scaleanchor
            If set to another axis id (e.g. `x2`, `y`), the range
            of this axis changes together with the range of the
            corresponding axis such that the scale of pixels per
            unit is in a constant ratio. Both axes are still
            zoomable, but when you zoom one, the other will zoom
            the same amount, keeping a fixed midpoint. `constrain`
            and `constraintoward` determine how we enforce the
            constraint. You can chain these, ie `yaxis:
            {scaleanchor: *x*}, xaxis2: {scaleanchor: *y*}` but you
            can only link axes of the same `type`. The linked axis
            can have the opposite letter (to constrain the aspect
            ratio) or the same letter (to match scales across
            subplots). Loops (`yaxis: {scaleanchor: *x*}, xaxis:
            {scaleanchor: *y*}` or longer) are redundant and the
            last constraint encountered will be ignored to avoid
            possible inconsistent constraints via `scaleratio`.
            Note that setting axes simultaneously in both a
            `scaleanchor` and a `matches` constraint is currently
            forbidden. Setting `false` allows to remove a default
            constraint (occasionally, you may need to prevent a
            default `scaleanchor` constraint from being applied,
            eg. when having an image trace `yaxis: {scaleanchor:
            "x"}` is set automatically in order for pixels to be
            rendered as squares, setting `yaxis: {scaleanchor:
            false}` allows to remove the constraint).
        scaleratio
            If this axis is linked to another by `scaleanchor`,
            this determines the pixel to unit scale ratio. For
            example, if this value is 10, then every unit on this
            axis spans 10 times the number of pixels as a unit on
            the linked axis. Use this for example to create an
            elevation profile where the vertical scale is
            exaggerated a fixed amount with respect to the
            horizontal.
        separatethousands
            If "true", even 4-digit integers are separated
        shift
            Moves the axis a given number of pixels from where it
            would have been otherwise. Accepts both positive and
            negative values, which will shift the axis either right
            or left, respectively. If `autoshift` is set to true,
            then this defaults to a padding of -3 if `side` is set
            to "left". and defaults to +3 if `side` is set to
            "right". Defaults to 0 if `autoshift` is set to false.
            Only has an effect if `anchor` is set to "free".
        showdividers
            Determines whether or not a dividers are drawn between
            the category levels of this axis. Only has an effect on
            "multicategory" axes.
        showexponent
            If "all", all exponents are shown besides their
            significands. If "first", only the exponent of the
            first tick is shown. If "last", only the exponent of
            the last tick is shown. If "none", no exponents appear.
        showgrid
            Determines whether or not grid lines are drawn. If
            True, the grid lines are drawn at every tick mark.
        showline
            Determines whether or not a line bounding this axis is
            drawn.
        showspikes
            Determines whether or not spikes (aka droplines) are
            drawn for this axis. Note: This only takes affect when
            hovermode = closest
        showticklabels
            Determines whether or not the tick labels are drawn.
        showtickprefix
            If "all", all tick labels are displayed with a prefix.
            If "first", only the first tick is displayed with a
            prefix. If "last", only the last tick is displayed with
            a suffix. If "none", tick prefixes are hidden.
        showticksuffix
            Same as `showtickprefix` but for tick suffixes.
        side
            Determines whether a x (y) axis is positioned at the
            "bottom" ("left") or "top" ("right") of the plotting
            area.
        spikecolor
            Sets the spike color. If undefined, will use the series
            color
        spikedash
            Sets the dash style of lines. Set to a dash type string
            ("solid", "dot", "dash", "longdash", "dashdot", or
            "longdashdot") or a dash length list in px (eg
            "5px,10px,2px,2px").
        spikemode
            Determines the drawing mode for the spike line If
            "toaxis", the line is drawn from the data point to the
            axis the  series is plotted on. If "across", the line
            is drawn across the entire plot area, and supercedes
            "toaxis". If "marker", then a marker dot is drawn on
            the axis the series is plotted on
        spikesnap
            Determines whether spikelines are stuck to the cursor
            or to the closest datapoints.
        spikethickness
            Sets the width (in px) of the zero line.
        tick0
            Sets the placement of the first tick on this axis. Use
            with `dtick`. If the axis `type` is "log", then you
            must take the log of your starting tick (e.g. to set
            the starting tick to 100, set the `tick0` to 2) except
            when `dtick`=*L<f>* (see `dtick` for more info). If the
            axis `type` is "date", it should be a date string, like
            date data. If the axis `type` is "category", it should
            be a number, using the scale where each category is
            assigned a serial number from zero in the order it
            appears.
        tickangle
            Sets the angle of the tick labels with respect to the
            horizontal. For example, a `tickangle` of -90 draws the
            tick labels vertically.
        tickcolor
            Sets the tick color.
        tickfont
            Sets the tick font.
        tickformat
            Sets the tick label formatting rule using d3 formatting
            mini-languages which are very similar to those in
            Python. For numbers, see:
            https://github.com/d3/d3-format/tree/v1.4.5#d3-format.
            And for dates see: https://github.com/d3/d3-time-
            format/tree/v2.2.3#locale_format. We add two items to
            d3's date formatter: "%h" for half of the year as a
            decimal number as well as "%{n}f" for fractional
            seconds with n digits. For example, *2016-10-13
            09:15:23.456* with tickformat "%H~%M~%S.%2f" would
            display "09~15~23.46"
        tickformatstops
            A tuple of :class:`plotly.graph_objects.layout.yaxis.Ti
            ckformatstop` instances or dicts with compatible
            properties
        tickformatstopdefaults
            When used in a template (as
            layout.template.layout.yaxis.tickformatstopdefaults),
            sets the default property values to use for elements of
            layout.yaxis.tickformatstops
        ticklabelindex
            Only for axes with `type` "date" or "linear". Instead
            of drawing the major tick label, draw the label for the
            minor tick that is n positions away from the major
            tick. E.g. to always draw the label for the minor tick
            before each major tick, choose `ticklabelindex` -1.
            This is useful for date axes with `ticklabelmode`
            "period" if you want to label the period that ends with
            each major tick instead of the period that begins
            there.
        ticklabelindexsrc
            Sets the source reference on Chart Studio Cloud for
            `ticklabelindex`.
        ticklabelmode
            Determines where tick labels are drawn with respect to
            their corresponding ticks and grid lines. Only has an
            effect for axes of `type` "date" When set to "period",
            tick labels are drawn in the middle of the period
            between ticks.
        ticklabeloverflow
            Determines how we handle tick labels that would
            overflow either the graph div or the domain of the
            axis. The default value for inside tick labels is *hide
            past domain*. Otherwise on "category" and
            "multicategory" axes the default is "allow". In other
            cases the default is *hide past div*.
        ticklabelposition
            Determines where tick labels are drawn with respect to
            the axis Please note that top or bottom has no effect
            on x axes or when `ticklabelmode` is set to "period".
            Similarly left or right has no effect on y axes or when
            `ticklabelmode` is set to "period". Has no effect on
            "multicategory" axes or when `tickson` is set to
            "boundaries". When used on axes linked by `matches` or
            `scaleanchor`, no extra padding for inside labels would
            be added by autorange, so that the scales could match.
        ticklabelshift
            Shifts the tick labels by the specified number of
            pixels in parallel to the axis. Positive values move
            the labels in the positive direction of the axis.
        ticklabelstandoff
            Sets the standoff distance (in px) between the axis
            tick labels and their default position. A positive
            `ticklabelstandoff` moves the labels farther away from
            the plot area if `ticklabelposition` is "outside", and
            deeper into the plot area if `ticklabelposition` is
            "inside". A negative `ticklabelstandoff` works in the
            opposite direction, moving outside ticks towards the
            plot area and inside ticks towards the outside. If the
            negative value is large enough, inside ticks can even
            end up outside and vice versa.
        ticklabelstep
            Sets the spacing between tick labels as compared to the
            spacing between ticks. A value of 1 (default) means
            each tick gets a label. A value of 2 means shows every
            2nd label. A larger value n means only every nth tick
            is labeled. `tick0` determines which labels are shown.
            Not implemented for axes with `type` "log" or
            "multicategory", or when `tickmode` is "array".
        ticklen
            Sets the tick length (in px).
        tickmode
            Sets the tick mode for this axis. If "auto", the number
            of ticks is set via `nticks`. If "linear", the
            placement of the ticks is determined by a starting
            position `tick0` and a tick step `dtick` ("linear" is
            the default value if `tick0` and `dtick` are provided).
            If "array", the placement of the ticks is set via
            `tickvals` and the tick text is `ticktext`. ("array" is
            the default value if `tickvals` is provided). If
            "sync", the number of ticks will sync with the
            overlayed axis set by `overlaying` property.
        tickprefix
            Sets a tick label prefix.
        ticks
            Determines whether ticks are drawn or not. If "", this
            axis' ticks are not drawn. If "outside" ("inside"),
            this axis' are drawn outside (inside) the axis lines.
        tickson
            Determines where ticks and grid lines are drawn with
            respect to their corresponding tick labels. Only has an
            effect for axes of `type` "category" or
            "multicategory". When set to "boundaries", ticks and
            grid lines are drawn half a category to the left/bottom
            of labels.
        ticksuffix
            Sets a tick label suffix.
        ticktext
            Sets the text displayed at the ticks position via
            `tickvals`. Only has an effect if `tickmode` is set to
            "array". Used with `tickvals`.
        ticktextsrc
            Sets the source reference on Chart Studio Cloud for
            `ticktext`.
        tickvals
            Sets the values at which ticks on this axis appear.
            Only has an effect if `tickmode` is set to "array".
            Used with `ticktext`.
        tickvalssrc
            Sets the source reference on Chart Studio Cloud for
            `tickvals`.
        tickwidth
            Sets the tick width (in px).
        title
            :class:`plotly.graph_objects.layout.yaxis.Title`
            instance or dict with compatible properties
        type
            Sets the axis type. By default, plotly attempts to
            determined the axis type by looking into the data of
            the traces that referenced the axis in question.
        uirevision
            Controls persistence of user-driven changes in axis
            `range`, `autorange`, and `title` if in `editable:
            true` configuration. Defaults to `layout.uirevision`.
        visible
            A single toggle to hide the axis while preserving
            interaction like dragging. Default is true when a
            cheater plot is present on the axis, otherwise false
        zeroline
            Determines whether or not a line is drawn at along the
            0 value of this axis. If True, the zero line is drawn
            on top of the grid lines.
        zerolinecolor
            Sets the line color of the zero line.
        zerolinewidth
            Sets the width (in px) of the zero line.
        
Did you mean "tickfont"?

Bad property path:
titlefont
^^^^^^^^^

Caption: This bar chart contrasts CO₂ emissions per capita (left axis, orange bars) with healthy life expectancy (right axis, blue bars) for three example countries in 2019. The United States emits far more CO₂ per person (≈15 tons) than France (≈5 tons) or India (<2 tons). Yet, Americans have a shorter healthy life expectancy (~66 years) than the French (~72 years). Indians have a much lower healthy lifespan (~60 years) alongside very low emissions.

Insight: The comparison highlights that more emissions do not guarantee better health. The U.S. vs France gap is telling: Americans emit about 3× more CO₂ per capita, but enjoy roughly 6 fewer healthy years on average than the French. This could be due to pollution or other lifestyle and healthcare differences, perspective 1 would note that high emissions (often accompanied by pollution and greenhouse effects) might be undermining health in the U.S., which struggles with issues like air quality and chronic disease. Meanwhile, India shows the opposite extreme: very low emissions come with low healthy life expectancy, primarily due to poverty and limited healthcare. While India’s low emissions are not causing poor health (rather, they reflect less industrial development), perspective 1 advocates worry that as India’s emissions rise, environmental health burdens (e.g. smog in cities) could further challenge its progress. This case study suggests that beyond a certain point, increasing emissions is associated with diminishing health returns, France achieves higher health with lower emissions than the U.S.,aligning with the idea that cleaner development paths might support longer healthy lives.

Visualization 3. Air Quality and Health Trend – China as a Case (2000–2021)#

– Does rapid emission growth slow health progress?

Hide code cell source
# ── Complete Cel: Plotly-plot met notebook_connected renderer ──

import pandas as pd
import plotly.graph_objects as go
import plotly.io as pio

# Zet renderer op notebook_connected voor inline weergave
pio.renderers.default = "notebook_connected"

# Data inladen & filteren
co2 = pd.read_csv('owid-co2-data.csv', usecols=['country','year','co2_per_capita'])
hale = pd.read_csv('C64284D_ALL_LATEST.csv')

# Harmoniseer & converteer jaartal
co2['year'] = co2['year'].astype(int)
hale = (
    hale.loc[
        (hale['IND_NAME']=='Healthy life expectancy (at birth)') &
        (hale['DIM_SEX']=='TOTAL'),
        ['GEO_NAME_SHORT','DIM_TIME','AMOUNT_N']
    ]
    .rename(columns={
        'GEO_NAME_SHORT':'country',
        'DIM_TIME':'year',
        'AMOUNT_N':'healthy_life_expectancy'
    })
)
hale['year'] = hale['year'].astype(int)

# Filter op China 2000–2021 en merge
co2_china  = co2.query("country=='China' and 2000 <= year <= 2021")
hale_china = hale.query("country=='China' and 2000 <= year <= 2021")
china = pd.merge(co2_china, hale_china, on=['country','year']).sort_values('year')

# Bouw figuur
fig = go.Figure()
fig.add_trace(go.Scatter(
    x=china['year'], y=china['co2_per_capita'],
    name='CO₂ per Capita (tons)', mode='lines+markers',
    marker=dict(color='orange'), line=dict(color='orange'),
    yaxis='y1', hovertemplate='CO₂: %{y:.2f} ton<extra></extra>'
))
fig.add_trace(go.Scatter(
    x=china['year'], y=china['healthy_life_expectancy'],
    name='Healthy Life Expectancy (years)', mode='lines+markers',
    marker=dict(color='blue'), line=dict(color='blue'),
    yaxis='y2', hovertemplate='Gezonde levensverwachting: %{y:.1f} jaar<extra></extra>'
))
fig.update_layout(
    title='China: Trend of CO₂ per Capita and Healthy Life Expectancy (2000–2021)',
    xaxis=dict(title='Year'),
    yaxis=dict(
        title='CO₂ per Capita (tons)',
        titlefont=dict(color='orange'),
        tickfont=dict(color='orange')
    ),
    yaxis2=dict(
        title='Healthy Life Expectancy (years)',
        titlefont=dict(color='blue'),
        tickfont=dict(color='blue'),
        overlaying='y', side='right'
    ),
    legend=dict(orientation='h', yanchor='bottom', y=1.02, x=0.5, xanchor='center'),
    margin=dict(t=80, b=50, l=60, r=60)
)

# Toon grafiek inline
fig.show()

Caption: This line chart shows China’s CO₂ emissions per capita (orange line, in tons) and healthy life expectancy at birth (blue line, in years) from 2000 to 2021. China’s CO₂ per person surged dramatically (more than tripling over two decades), while healthy life expectancy also rose steadily (from about 62 up to ~68 years).

Insight: China illustrates a nuanced story. Despite severe pollution challenges during its rapid industrialization, healthy life expectancy still improved significantly as the country became wealthier. There isn’t an obvious dip or slowdown in the upward health trend corresponding to rising emissions in fact, both lines climb upward. Advocates of perspective 1, however, argue that China’s health gains could have been even greater without the heavy air pollution that accompanied its coal-driven economic boom. In the 2010s, recognizing these issues, China enacted aggressive clean air policies. Noticeably, the curve of CO₂ per capita leveled off slightly toward 2019, and pollution levels in cities started dropping, which may help future health outcomes. This example underscores that while high emissions haven’t stopped health improvements outright, they likely impose hidden costs: respiratory illnesses, environmental stress, and fewer healthy years than might be possible in a cleaner environment. The full negative impact of emissions might be long-term through climate change and not fully captured in this 20-year window. Overall, the data for perspective 1 shows some hints (like the U.S. vs France comparison) that excessive emissions and pollution correlate with health drawbacks, but the relationship is complex and often outweighed by socioeconomic factors.

Perspective 2: Are Other Factors More Important than Emissions for Life Expectancy?#

Perspective 2 suggests that development and public health infrastructure drive healthy lifespan, not emissions per se. If this is true, we expect to see that as countries get wealthier (often accompanied by more emissions), their healthy life expectancy increases – indicating a positive or neutral relationship between emissions and health when development is accounted for. We also expect that countries with the longest healthy lifespans are those with strong healthcare and high living standards, rather than the lowest emitters. The following visuals explore these patterns.

Visualization 4. Historical CO₂ Emissions vs Healthy Life Expectancy#

– Do countries that industrialized (high historic emissions) have longer healthy lives?

Hide code cell source
import plotly.express as px

# World map with CO2 per capita and Healthy Life Expectancy
fig = px.scatter_geo(
    merged_df[merged_df['year'] == 2019],
    locations='iso_code',
    color='healthy_life_expectancy_at_birth',
    size='co2_per_capita',
    color_continuous_scale='Viridis',
    hover_name='country',
    title='Healthy Life Expectancy (color) and CO₂ Emissions per Capita (size) in 2019'
)

fig.show()
---------------------------------------------------------------------------
ValueError                                Traceback (most recent call last)
Cell In[30], line 4
      1 import plotly.express as px
      3 # World map with CO2 per capita and Healthy Life Expectancy
----> 4 fig = px.scatter_geo(
      5     merged_df[merged_df['year'] == 2019],
      6     locations='iso_code',
      7     color='healthy_life_expectancy_at_birth',
      8     size='co2_per_capita',
      9     color_continuous_scale='Viridis',
     10     hover_name='country',
     11     title='Healthy Life Expectancy (color) and CO₂ Emissions per Capita (size) in 2019'
     12 )
     14 fig.show()

File c:\Users\haldo\anaconda3\Lib\site-packages\plotly\express\_chart_types.py:1148, in scatter_geo(data_frame, lat, lon, locations, locationmode, geojson, featureidkey, color, text, symbol, facet_row, facet_col, facet_col_wrap, facet_row_spacing, facet_col_spacing, hover_name, hover_data, custom_data, size, animation_frame, animation_group, category_orders, labels, color_discrete_sequence, color_discrete_map, color_continuous_scale, range_color, color_continuous_midpoint, symbol_sequence, symbol_map, opacity, size_max, projection, scope, center, fitbounds, basemap_visible, title, template, width, height)
   1101 def scatter_geo(
   1102     data_frame=None,
   1103     lat=None,
   (...)
   1142     height=None,
   1143 ) -> go.Figure:
   1144     """
   1145     In a geographic scatter plot, each row of `data_frame` is represented
   1146     by a symbol mark on a map.
   1147     """
-> 1148     return make_figure(
   1149         args=locals(),
   1150         constructor=go.Scattergeo,
   1151         trace_patch=dict(locationmode=locationmode),
   1152     )

File c:\Users\haldo\anaconda3\Lib\site-packages\plotly\express\_core.py:2117, in make_figure(args, constructor, trace_patch, layout_patch)
   2114 layout_patch = layout_patch or {}
   2115 apply_default_cascade(args)
-> 2117 args = build_dataframe(args, constructor)
   2118 if constructor in [go.Treemap, go.Sunburst, go.Icicle] and args["path"] is not None:
   2119     args = process_dataframe_hierarchy(args)

File c:\Users\haldo\anaconda3\Lib\site-packages\plotly\express\_core.py:1513, in build_dataframe(args, constructor)
   1510     args["color"] = None
   1511 # now that things have been prepped, we do the systematic rewriting of `args`
-> 1513 df_output, wide_id_vars = process_args_into_dataframe(
   1514     args, wide_mode, var_name, value_name
   1515 )
   1517 # now that `df_output` exists and `args` contains only references, we complete
   1518 # the special-case and wide-mode handling by further rewriting args and/or mutating
   1519 # df_output
   1521 count_name = _escape_col_name(df_output, "count", [var_name, value_name])

File c:\Users\haldo\anaconda3\Lib\site-packages\plotly\express\_core.py:1234, in process_args_into_dataframe(args, wide_mode, var_name, value_name)
   1232         if argument == "index":
   1233             err_msg += "\n To use the index, pass it in directly as `df.index`."
-> 1234         raise ValueError(err_msg)
   1235 elif length and len(df_input[argument]) != length:
   1236     raise ValueError(
   1237         "All arguments should have the same length. "
   1238         "The length of column argument `df[%s]` is %d, whereas the "
   (...)
   1245         )
   1246     )

ValueError: Value of 'locations' is not the name of a column in 'data_frame'. Expected one of ['country', 'year', 'co2_per_capita', 'healthy_life_expectancy_at_birth'] but received: iso_code

Caption: Each point is a country (2019), plotting total historical CO₂ emitted (on a log scale) against healthy life expectancy. A clear upward trend emerges: countries with a larger cumulative CO₂ footprint (toward the right) tend to have higher healthy life expectancy.

Insight: This chart reveals a strong positive association: nations that have emitted the most CO₂ over history, typically more developed economies, almost all enjoy high healthy life expectancies (70+ years). For instance, countries like Japan, Germany, the UK, or the U.S. (far right) have among the longest healthy lifespans. Conversely, countries with negligible historical emissions (far left) are generally those with shorter healthy lives (often under 60 years). This does not mean emitting CO₂ causes people to live longer; rather, it indicates that industrialization and development, which inevitably came with CO₂ emissions, enabled better health outcomes. In other words, wealth and infrastructure correlate with both high emissions and high life expectancy. This aligns with Perspective 2: healthy longevity is achieved through improved hospitals, nutrition, education, and living conditions which have historically been financed by the economic growth that also drove up emissions. The implication is that life expectancy can rise alongside emissions if development is occurring, and that cutting emissions need not reduce life expectancy as long as development and healthcare are maintained.

Visualization 5. Emissions vs Healthy Life Expectancy Over Time (Interactive)#

– How have emissions and health evolved together from 2000 to 2021?

Hide code cell source
import pandas as pd
import matplotlib.pyplot as plt
import numpy as np
from matplotlib.widgets import Slider

# 1) Data inladen en voorbereiden zoals je al deed
co2 = pd.read_csv('owid-co2-data.csv', usecols=['country','year','co2_per_capita'])
hale = pd.read_csv('C64284D_ALL_LATEST.csv')

hale = (
    hale.loc[
        (hale['IND_NAME']=='Healthy life expectancy (at birth)') &
        (hale['DIM_SEX']=='TOTAL'),
        ['GEO_NAME_SHORT','DIM_TIME','AMOUNT_N']
    ]
    .rename(columns={
        'GEO_NAME_SHORT':'country',
        'DIM_TIME':'year',
        'AMOUNT_N':'healthy_life_expectancy'
    })
)

co2['year']  = co2['year'].astype(int)
hale['year'] = hale['year'].astype(int)

df = pd.merge(co2, hale, on=['country','year'], how='inner')
df = df[(df.year >= 2000) & (df.year <= 2021)]
years = sorted(df['year'].unique())

# 2) Figuur en slider-as opzetten
fig, ax = plt.subplots(figsize=(8,6))
plt.subplots_adjust(bottom=0.15)  # ruimte voor de slider

# Initiële scatter voor het eerste jaar
current_year = years[0]
sub = df[df.year == current_year]
scat = ax.scatter(sub['co2_per_capita'], sub['healthy_life_expectancy'],
                  s=50, edgecolor='k')
title = ax.set_title(f'Year: {current_year}')

ax.set_xlim(0, df['co2_per_capita'].max()*1.1)
ax.set_ylim(df['healthy_life_expectancy'].min()*0.9,
            df['healthy_life_expectancy'].max()*1.05)
ax.set_xlabel('CO₂ per Capita (t/person)')
ax.set_ylabel('Healthy Life Expectancy (years)')
ax.grid(True, linestyle='--', alpha=0.5)

# Slider-as: [left, bottom, width, height] in fraction of fig
ax_slider = fig.add_axes([0.15, 0.05, 0.7, 0.03])
slider = Slider(
    ax=ax_slider,
    label='Year',
    valmin=years[0],
    valmax=years[-1],
    valinit=current_year,
    valstep=years,
    color='lightblue'
)

# 3) Update-functie voor de slider
def update(val):
    yr = int(slider.val)
    sub = df[df.year == yr]
    scat.set_offsets(np.c_[sub['co2_per_capita'], sub['healthy_life_expectancy']])
    title.set_text(f'Year: {yr}')
    fig.canvas.draw_idle()

slider.on_changed(update)

plt.show()
_images/e30f7a94113c7c658742cb0f129f97ab13feab394754b45006471cda3a8033f1.png

Caption: This interactive bubble chart (select the play button) shows countries moving from 2000 to 2021. Bubbles typically drift upwards and to the right over time, meaning both CO₂ emissions per capita and healthy life expectancy have increased in tandem for many countries. Insight: The animation reinforces that in the last two decades, life expectancy improvements often coincided with rising emissions. Developing countries (with lower starting health and emissions) move markedly up-right: for example, India and Bangladesh start near the bottom-left in 2000 (low emissions, ~50s HALE) and progress upward by 2019 (somewhat higher emissions, HALE in 60s). China’s bubble shoots to the right (big emission surge) and also climbs upward (HALE from low 60s to high 60s). Most high-income countries were already in the upper-right and tend to inch further up (health gains) even as their emissions per capita plateau or decline slightly. Notably, European countries have modest or falling CO₂ per capita but still improve health to around 70+ healthy years, showing that it’s possible to gain in health while curbing emissions. The overall picture supports Perspective 2: there is no general trade-off where increasing emissions universally lowers life expectancy if anything, countries have managed to raise healthy life expectancy substantially despite higher emissions. This suggests that improving medical care, sanitation, education, and incomes (which often come with industrial growth) has a more immediate and powerful effect on health than the hypothesized negative effects of CO₂. Of course, this does not mean CO₂-driven climate change has no future health impact; rather, up to 2021, socioeconomic progress appears to outweigh any direct life expectancy harms from emissions.

Visualization 6. Global Overview: Emissions and Health by Country (2019)#

– Who has high emissions, and who lives long and healthy?

# 1. Libraries importeren en renderer instellen
import pandas as pd
import plotly.express as px
import plotly.io as pio

# Kies de juiste renderer: 
# - klassieke Notebook: 'notebook' 
# - JupyterLab:      'jupyterlab'
pio.renderers.default = 'notebook'

# 2. WHO Healthy Life Expectancy inladen & filteren
who_df = pd.read_csv('C64284D_ALL_LATEST.csv')
who_df = who_df[
    (who_df['IND_NAME'] == 'Healthy life expectancy (at birth)') &
    (who_df['DIM_GEO_CODE_TYPE'] == 'COUNTRY')
]
who_df = who_df.rename(columns={
    'GEO_NAME_SHORT': 'country',
    'DIM_TIME':        'year',
    'AMOUNT_N':        'healthy_life_expectancy_at_birth'
})[['country','year','healthy_life_expectancy_at_birth']]
who_df['year'] = who_df['year'].astype(int)

# 3. OWID CO₂ per capita inladen & selecteren
co2_df = pd.read_csv('owid-co2-data.csv')
co2_df = co2_df[['country','year','co2_per_capita']]
co2_df['year'] = co2_df['year'].astype(int)

# 4. Beide datasets mergen
merged_df = pd.merge(
    co2_df,
    who_df,
    how='inner',
    on=['country','year']
)

# 5. Direct inline tonen van de interactieve animatie
fig = px.scatter(
    merged_df,
    x='co2_per_capita',
    y='healthy_life_expectancy_at_birth',
    animation_frame='year',
    animation_group='country',
    hover_name='country',
    range_x=[0,35],
    range_y=[40,85],
    labels={
        'co2_per_capita':                    'CO₂ per Capita (t/person)',
        'healthy_life_expectancy_at_birth': 'Healthy Life Expectancy (years)'
    },
    title='Emissions vs Healthy Life Expectancy (2000–2021)'
)
fig.show()

Caption: In this world map, bubble size represents CO₂ emissions per capita and color represents healthy life expectancy (yellow-green = shorter healthy lives, blue-purple = longer healthy lives). We see large bubbles concentrated in North America, the Middle East, and parts of Asia/Oceania (indicating high per-person emissions), whereas small bubbles cover most of Africa and South Asia (minimal emissions). Crucially, many large bubbles are colored blue/purple – for example, the US, Canada, Australia, and Gulf states have high HALE (~65–75 years) despite high emissions. In contrast, the smallest bubbles (low emitters) in sub-Saharan Africa are often yellowish, showing low healthy life expectancy (50s).

Insight: This global view underscores that long healthy lives are achieved across a range of emission levels, and the worst health outcomes are mostly in low-emission, low-income countries. High emissions per capita are primarily a feature of wealthy nations and oil producers. These countries generally have the infrastructure for high life expectancy (though not always the very highest: e.g., the USA’s HALE is lower than some lower-emission European countries). Meanwhile, countries with the shortest healthy lifespans are poor and emit very little CO₂. This pattern supports the idea that economic and health system development are the dominant factors for healthy life expectancy, not CO₂ levels. If anything, the map suggests an injustice: those who have contributed least to emissions (and climate change) often have the lowest life expectancies. It also implies that reducing emissions in high-HALE countries (for sustainability) should be feasible without sacrificing their hard-won health outcomes, given their strong health systems. In summary, Perspective 2 finds that life expectancy is more strongly tied to wealth and public health investments than to carbon emissions. Policies focusing on improving healthcare, nutrition, and the environment together could continue to raise healthy life expectancy while also cutting unnecessary CO₂ emissions.

References:#